similar to: [PATCH 1/2] fb/ramgf100: disable memory reclocking by default

Todo: - Determine source of R[10f298] & 0x11 Signed-off-by: Roy Spliet <nouveau at spliet.org> --- drivers/gpu/drm/nouveau/nvkm/subdev/fb/ramgf100.c | 115 ++++++++++++++-------- 1 file changed, 76 insertions(+), 39 deletions(-) diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/fb/ramgf100.c b/drivers/gpu/drm/nouveau/nvkm/subdev/fb/ramgf100.c index fffd01a..6ebdc4c 100644 ---

some chipset have working engine reclocking, but broken memory reclocking like Fermi. We should for now, add the functionality to allow partly reclocking for those. Allthough this doesn't give as much performance as one might wish, it is till noticeable and may improve performance enough to be noted. Karol Herbst (2): clk: seperate engine and memory reclock toggles clk: allow engine

I have looked at problem with Fermi GPUs where changing to higher clock led to really bad perfomance (with GpuTest 20x worse perfomance) and later also crashes of the nouveau. It seemed to be affected by Shader Clock in Voltage Entries in the video BIOS. Disabling BYPASS_PLL_CHECK in CLK0_CTRL seems to completely fix the issue. I have tried to search this BYPASS_PLL_CHECK in Nvidia traces but

Parts are re-used even on NVA3, others from GF100 on Signed-off-by: Roy Spliet <nouveau at spliet.org> --- drivers/gpu/drm/nouveau/nvkm/subdev/fb/ram.h | 17 +++ drivers/gpu/drm/nouveau/nvkm/subdev/fb/ramgf100.c | 92 +++++++++----- drivers/gpu/drm/nouveau/nvkm/subdev/fb/ramgk104.c | 140 +--------------------- drivers/gpu/drm/nouveau/nvkm/subdev/fb/ramgt215.c | 61 ++++++++++ 4

Probably want ULL for 32-bit arches to be correct here too. On Tue, Dec 31, 2019 at 3:53 PM Wambui Karuga <wambui.karugax at gmail.com> wrote: > > Explicitly declare constants are unsigned long to address the following > sparse warnings: > warning: constant is so big it is long > > Signed-off-by: Wambui Karuga <wambui.karugax at gmail.com> > --- >

Explicitly declare constants as unsigned long long to address the following sparse warnings: warning: constant is so big it is long v2: convert to unsigned long long for compatibility with 32-bit architectures. Signed-off-by: Wambui Karuga <wambui.karugax at gmail.com> Suggested by: lia Mirkin <imirkin at alum.mit.edu> --- drivers/gpu/drm/nouveau/nvkm/subdev/fb/ramgf100.c | 2 +-

Explicitly declare constants are unsigned long to address the following sparse warnings: warning: constant is so big it is long Signed-off-by: Wambui Karuga <wambui.karugax at gmail.com> --- drivers/gpu/drm/nouveau/nvkm/subdev/fb/ramgf100.c | 2 +- drivers/gpu/drm/nouveau/nvkm/subdev/fb/ramgf108.c | 2 +- drivers/gpu/drm/nouveau/nvkm/subdev/fb/ramgk104.c | 2 +-

Signed-off by: Rosen Penev <rosenp at gmail.com> --- drm/nouveau/nvkm/subdev/fb/ramgf100.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drm/nouveau/nvkm/subdev/fb/ramgf100.c b/drm/nouveau/nvkm/subdev/fb/ramgf100.c index 4a9bd4f1..feb8820c 100644 --- a/drm/nouveau/nvkm/subdev/fb/ramgf100.c +++ b/drm/nouveau/nvkm/subdev/fb/ramgf100.c @@ -717,7 +717,7 @@

This is fixed in a different nouveau branch. from Ben Skeggs last I remember. On Fri, Dec 29, 2017 at 1:37 PM, Pierre Moreau <pierre.morrow at free.fr> wrote: > The typo is still present in master. > Not sure whether it will help you Ben, regarding Fermi reclocking, but it > shouldn’t harm in reducing possible errors. :-) > > Regards, > Pierre > > On 2017-07-27 —

Just because the blob does it too... Signed-off-by: Roy Spliet <nouveau at spliet.org> --- drivers/gpu/drm/nouveau/nvkm/subdev/fb/ramgf100.c | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/fb/ramgf100.c b/drivers/gpu/drm/nouveau/nvkm/subdev/fb/ramgf100.c index eef09bf..62359c2 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/fb/ramgf100.c +++

It's not found on Fermi. Signed-off-by: Roy Spliet <nouveau at spliet.org> --- drivers/gpu/drm/nouveau/nvkm/subdev/fb/ramgf100.c | 13 ++++++++----- 1 file changed, 8 insertions(+), 5 deletions(-) diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/fb/ramgf100.c b/drivers/gpu/drm/nouveau/nvkm/subdev/fb/ramgf100.c index 38a7e2b..eef09bf 100644 ---

Signed-off-by: Roy Spliet <nouveau at spliet.org> --- drivers/gpu/drm/nouveau/nvkm/subdev/fb/ramgf100.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/fb/ramgf100.c b/drivers/gpu/drm/nouveau/nvkm/subdev/fb/ramgf100.c index a469719..eebd20b 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/fb/ramgf100.c +++

Signed-off-by: Roy Spliet <nouveau at spliet.org> --- drivers/gpu/drm/nouveau/nvkm/subdev/fb/ramgf100.c | 59 +++++++++++++++++------ 1 file changed, 43 insertions(+), 16 deletions(-) diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/fb/ramgf100.c b/drivers/gpu/drm/nouveau/nvkm/subdev/fb/ramgf100.c index 62359c2..a469719 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/fb/ramgf100.c +++

The typo is still present in master. Not sure whether it will help you Ben, regarding Fermi reclocking, but it shouldn’t harm in reducing possible errors. :-) Regards, Pierre On 2017-07-27 — 18:20, Rosen Penev wrote: > Signed-off by: Rosen Penev <rosenp at gmail.com> > --- > drm/nouveau/nvkm/subdev/fb/ramgf100.c | 2 +- > 1 file changed, 1 insertion(+), 1 deletion(-) >

In correspondence with the other ram_calc implementations. Signed-off-by: Roy Spliet <nouveau at spliet.org> --- drivers/gpu/drm/nouveau/nvkm/subdev/fb/ramgf100.c | 37 ++++++++++++----------- 1 file changed, 19 insertions(+), 18 deletions(-) diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/fb/ramgf100.c b/drivers/gpu/drm/nouveau/nvkm/subdev/fb/ramgf100.c index 093223d..fffd01a 100644 ---

Hi Karol, I have some comments below. On 05:42 PM - Dec 01 2015, Karol Herbst wrote: > 0: disable boosting (cap to base clock from the vbios) > 1: boost only to boost clock from the vbios > 2: boost to max clock available > --- > drm/nouveau/include/nvkm/subdev/clk.h | 10 +++++++++- > drm/nouveau/nvkm/subdev/clk/base.c | 17 ++++++++++++++++- >

this series solves different issues we encounter on kepler cards while reclocking: 1. core clock doesn't change at all and produces a volting error (patch 1) this can happen when the voltage table has only 0ed values in the header so we have to parse the entries itself, which contain the right voltages 2. kepler won't clock to highest cstates (patch 2) this happens, because there are

A missing u64 cast causes a 32-Bit wraparound from 4096 MiB to 0 MiB and therefore total 0 MiB VRAM detected if card has 4096 Mib per FBP. Signed-off-by: Mario Kleiner <mario.kleiner.de at gmail.com> --- drivers/gpu/drm/nouveau/nvkm/subdev/fb/ramgf100.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/fb/ramgf100.c

Some newer chips have trouble coming up, and we get bad MMIO reads from them, like 0xbadf100. This ends up translating into crazy amounts of VRAM, which destroys all sorts of other logic down the line. Instead, fail device init. Signed-off-by: Ilia Mirkin <imirkin at alum.mit.edu> Cc: stable at kernel.org --- drm/nouveau/nvkm/subdev/fb/ramgf100.c | 6 ++++++ 1 file changed, 6 insertions(+)

Signed-off-by: Roy Spliet <nouveau at spliet.org> --- drivers/gpu/drm/nouveau/include/nvkm/subdev/bios/perf.h | 1 + drivers/gpu/drm/nouveau/nvkm/subdev/bios/perf.c | 14 ++++++++++++++ drivers/gpu/drm/nouveau/nvkm/subdev/fb/ramgf100.c | 1 + 3 files changed, 16 insertions(+) diff --git a/drivers/gpu/drm/nouveau/include/nvkm/subdev/bios/perf.h